Effects of a False Neurotransmitter, p-Hydroxynorephedrine, on the Function ofAdrenergic Neurons in Hypertensive Patients

Robert E. Rangno, … , J. Throck Watson, John Oates

J Clin Invest. 1973;52(4):952-960. https://doi.org/10.1172/JCI107260.

Previous studies have shown that amphetamine and p-hydroxyamphetamine impairadrenergic transmission, and it has been suggested that this effect is mediated by an activemetabolite, p-hydroxynorephedrine (PHN). Studies in experimental animals have shownthat PHN can deplete and substitute for norepinephrine (NE) in the transmitter pool, thusmeeting the criteria of a false neurotransmitter.

The pharmacologic effects of PHN on adrenergic function and NE synthesis were studied ineight hypertensive patients and compared with placebo. Mean erect and supine bloodpressure (BP) decreased 22/14 and 9/6 mm Hg, respectively, during PHN 600 mg daily.The post-Valsalva diastolic overshoot was abolished. The pressor sensitivity to tyraminedecreased whereas the pressor response to NE was enhanced. A mild natriuresis occurred.The 24 h urinary excretion of catecholamines and catecholamine metabolites during theadministration of PHN compared with placebo changed as follows: vanillylmandelic acid(VMA), 42% decrease; NE. 42% decrease; normetanephrine (NM), 400% increase:metanephrine, unchanged; dopamine, 40% decrease; while homovanillic acid wasunchanged. The sum of VMA, NE, and NM decreased 23%. The posttreatment urinaryexcretion of PHN was biexponential with first and second phase half-lives of 13 and 55 h.respectively. The time of the second phase closely approximated the recovery of thechanges in BP and excretion of VMA. No effects of PHN on the central nervous systemwere observed. […]

Research Article

Find the latest version:

http://jci.me/107260-pdf

Effects of a False Neurotransmitter, p-Hydroxynorephedrine, on

the Function of Adrenergic Neurons in Hypertensive Patients

ROBERTE. RANGNO,JOHNS. KAUFMANN,JOHNH. CAVANAUGH,DONALDISLAND,

J. THRmcKWATSON,and JOHNOATES

From the Division of Clinical Pharmacology, Departments of Medicine andPharmacology, Vanderbilt University School of Medicine,Nashville, Tennessee 37203

A D S T R A C T Previous studies have shown that am-phetamine and p-hydroxyamphetamine impair adrenergictransmission, and it has been suggested that this ef-fect is mediated by an active metabolite, p-hydroxy-norephedrine (PHN). Studies in experimental ani-mals have shown that PHN can deplete and substitutefor norepinephrine (NE) in the transmitter pool, thusmeeting the criteria of a false neurotransmitter.

The pharmacologic effects of PHN on adrenergicfunction and NE synthesis were studied in eight hyper-tensive patients and compared with placebo. Mean erectand supine blood pressure (BP) decreased 22/14 and 9/6mmHg, respectively, during PHN 600 mg daily. Thepost-Valsalva diastolic overshoot was abolished. Thepressor sensitivity to tyramine decreased whereas thepressor response to NE was enhanced. A mild natriure-sis occurred. The 24 h urinary excretion of catechola-mines and catecholamine metabolites during the ad-ministration of PHNcompared with placebo changed asfollows: vanillylmandelic acid (VMA), 42% decrease;NE, 42% decrease; normetanephrine (NM), 400% in-crease; metanephrine, unchanged; dopamine, 40% de-crease; while homovanillic acid was unchanged. Thesum of VMA, NE, and NMdecreased 23%. The post-treatment urinary excretion of PHN was biexponentialwith first and second phase half-lives of 13 and 55 h, re-spectively. The time of the second phase closely ap-proximated the recovery of the changes in BP and ex-

This work was presented in part at the Annual Meetingof the American Federation for Clinical Research, AtlanticCity, N. J., May 1970.

Dr. Rangno was a fellow of the Canadian Foundationfor the Advancement of Therapeutics. His present address isthe Division of Clinical Pharmacology, Montreal GeneralHospital, Montreal 109, Quebec, Canada.

Received for publication 2 March 1972 and in revisedform 29 November 1972.

cretion of VMA. No effects of PHN on the centralnervous system were observed. These studies show thatPHN acts peripherally to interfere with adrenergicfunction and NE synthesis in hypertensive patientswith a resultant decrease in BP.

INTRODUCTIONRecent studies in this laboratory (1) have shown thatamphetamine is metabolized to p-hydroxynorephedrine(PHN)' [d,l-a- ( 1-aminoethyl) -p-hydroxybenzylalcoholerythro configuration] in man, presumably by the path-way depicted in Fig. 1. The known ability of PHN todeplete norephinephrine (NE) (2) and to replace it inthe neurosecetory stores (3) suggests that it may bean active metabolite which could interfere with adrener-gic neuron function by acting as a false neurotransmitter.This possibility is given credence by evidence of im-paired adrenergic neuron function after moderate tolarge doses of amphetamine (1). Also, p-hydroxyamphe-tamine (PHA), the intermediary metabolite in theconversion of amphetamine to PHN, has similarly beenshown to suppress reflex sympathetic function in normo-tensive human (4) and to lower the blood pressure(BP) in hypertensive dogs (5).

The present study was initiated to determine whetherPHN inhibited adrenergic function in man and to evalu-ate the possibility that introduction of this foreign aminemight interfere with NE biosynthesis.

1 Abbreviations used in this paper: BP, blood pressure;DA, dopamine; ECG, electrocardiogram; GLC-EC, gas-liquid chromatography with electron capture; HVA, homo-vanillic acid; M, metanephrine; NE, norepinephrine; NM,normetanephrine; PFP, pentafluoroproprionate; PHA, p-hy-droxyamphetamine; PHN, p-hydroxynorephedrine; RBBB,right bundle branch block; VMA, vanillylmandelic acid.

952 The Journal of Clinical Investigation Volume 52 April 1973

METHODS

Designi. Acute and subacute animal toxicity studies wereperformed initially and revealed no untoward effects ofPHN within the dose range employed in this study (6).Clinical studies were conducted in the Vanderbilt ClinicalResearch Center after a complete investigation of the pa-tients for hypertension and after their informed consent hadbeen obtained. The study included eight patients, four maleand four female, with a mean age of 41 yr (range: 36-59).The degree of hypertension ranged from moderate to severe.All patients had essential hypertension except for one pa-tient (M. L.) with diabetes mellitus who had remainedhypertensive after a nephrectomy for a unilateral renal ar-tery stenosis. All antihypertensive medications had beendiscontinued at least 7 days in advance, and a 100 meqsodium diet was started on admission (one patient requireda 10 meq sodium diet). The design included a single blindplacebo period followed by PHN (d, l-erythro-p-hydroxy-norephedrine) in oral, increasing, divided doses and a fol-low-up placebo period. Dose-ranging studies had been com-pleted previously in three other patients. Doses were givenevery 6 h, and occasionally every 4 h. BP was measured inthe supine and erect positions four time daily. BP wasrecorded three times at i-h intervals after any dose incre-ment. Urine was collected daily in 10 ml of 10 N HCl fordetermination of sodium, creatinine, catecholamines, freecatechol metabolites, and PHN. Urine samples were frozenbefore the analyses, except for NE determinations whichwere performed with fresh urine. Periodic assessment ofroutine hematologic, renal, and hepatic function was madeinitially, at the end of each dose increment, and after PHNwas discontinued.

Pharmtacologic tests. At the end of the initial placeboperiod, at the end of each effective PHN dose increment,and during the posttreatment placebo period, the followingstudies were performed: determination of the pressor sensi-tivity to the directly acting amine, norepinephrine, and tothe indirectly acting amine, tyramine; determination of thedepressor sensitivity to the alpha-blocking agent, phentol-amine; and the Valsalva maneuver. The supine patient wasprepared as follows: electrocardiogram(ECG)-monitoringleads were attached, indwelling intravenous and intra-ar-terial needles were placed, and a rest period of 15 minwas allowed before the above studies. Test medicationswere given as single, rapid i-v. doses with sufficient timebetween doses to allow measurements to return to base-linevalues.

The dose of NE and tyramine required to raise the sys-tolic pressure by 25 mm Hg was determined from thepartial dose-response curve. At least three dose-responsepoints were obtained for each test. The control dose wasdivided by that obtained during PHNadministration, result-ing in a ratio used to express the sensitivity to the pressoramine. The dose of phentolamine required to decrease thesystolic pressure by 15 mmHg was similarly determined.The ratio of control dose to treatment dose was used inthis case to express the depressor sensitivity to phentolamine.

The Valsalva maneuver was performed by the supinepatient maintaining a forced expiratory pressure of 40 mmHg for 10 sec (7). The test was repeated at least two timeson each occasion. The change in the pulse pressure duringforced expiration was calculated as a percentage decreaseof pretest pulse pressure. The change in the diastolic pres-sure in the immediate post-Valsalva period was calculatedas the percentage increase of the pretest diastolic pressure.

Urinary analyses. (a) Vanillylmandelic acid (VMA)

CH2- CH-NH2 amphetamine

I+0 (liver)

~~7~~~CHI3OH CHCHCH-NH2 p- OH-amphetamine

+0 (odrenergic vesicle)

{),OH- ~O CH3 ,p-OH norephedrine(PHN)

FIGURE 1 Biotransformation of amphetamine.

was measured spectrophotometrically after its conversionto vanillin (8). The presence of large concentrations ofPHN in the urine did not interfere with this assay.

(b) Free NE was measured fluorometrically (9). Thepresence of PHN in the urine caused a slight quench inthe assay which did not exceed 20%. Correction for quenchwas made by the use of internal standards.

(c) Normetanephrine (NM) and metanephrine (M) weremeasured by a modification of the assay of Anton andSayre (10). This assay was altered to include a furtherpaper chromatographic separation of NM and M fromPHN since large concentrations of PHN gave both quenchand intrinsic fluorescence. The overall recovery throughthis procedure, determined in every sample by the recoveryof tracer amounts of [3H]NMN and [3H]MN, was about50%. The specificity of the separation procedure was veri-fied by gas-liquid chromatography with electron capture(GLC-EC) plus mass spectrometry of a pentafluoropro-prionate (PFP) derivative of the NM sample.

(d) Dopamine (DA) was measured fluorometrically(11). The final eluate from this procedure was assayed forpossible interfering PHNby PFP derivatization and analy-sis on GLC-EC. The sensitivity of this method was 50 pgof PHN and verified the absence of PHN in the finaleluate.

(e) Homovanillic acid (HVA) was measured fluoro-metrically (12).

(f) PHN was measured spectrophotometrically after itsconversion to p-hydroxybenzaldehyde (13). Sensitivity ofthe assay was 0.5 gg/ml. After acid hydrolysis the samplewas adsorbed on an Amberlite CG-50 resin column (Roh-mand Haas Co., Philadelphia, Pa.). The column was elutedwith 4 N NH40H. A 4 ml sample of the eluate was oxi-dized with sodium metaperiodate and the UV absorptionmeasured at 333 nm. Non-acid-hydrolyzed samples wereassayed to determine the unconjugated (unchanged) portionof the total urinary PHN. Qualitative and quantitativeverification of some samples was made with GLC-EC analy-sis of the PFP derivative of PHN.

Statistical analysis. All statistical analysis of data wasperformed using the Student's t test.

RESULTSThe effect of PHNon BP. Systolic BPs in the erect

position decreased significantly in all patients duringthe administration of PHN 600 mg daily and the de-

Effects of p-Hydroxynorephedrine in Hypertensive Patients 953

TABLE 1

BPs in the Erect Position (mm Hg)

PHN

Patient Placebo 600 900 Placebo

mg/dayC. H. 149/100 129*/83* 148/102M. L. 149/108 116*/87* 141/107R. W'. 138/103 114*/88* 137/98D. K. 160/96 143*/97 149*/98 175/960. D. 150/107 137*/106 135*/97* 149/119F. P. 175,/141 135*/108* 130*/101* 176/132B. D. 151/109 137*/99* 149/108 165/116J. S. 163/120 146*/107* 136*/95* 164/114Mean decrease

from placebo 22.3/13.8 20.0/15.2

All values are means of four measurements daily for at least4 consecutive days.,* Differences between placebo and PHNsignificant P < 0.05with majority P << 0.01.

DAILY 900-DOSE 60s]PHN 300- PLACEBO PLACEBO

160 1

aBP 140 ;ok a

120 /

10°1i3o0 -T~ /I T -T --T-a-r -T-r> T-7II-BT -rT I

2 4 6 8 10BP14 16 18 20 22 24

120A~~~~~~~~~ A

DAYS

FIGURE 2 BP change in one patient (J. S.) during PHN.The erect systolic and diastolic pressures decreased sig-nificantly (P < 0.01) during PHN 600 mg daily with afurther decrease during PHN 900 mg daily.

crease in diastolic pressures was significant in seven ofthe eight patients. The mean decrease was 22.3/12.8 mmHg compared with placebo (Table I). The pressure ofthree patients fell to normotensive levels on PHN 600mg daily and this was their maximum dose. The dosewas increased in the remaining five patients to 900 mgdaily with little further change in supine and erect pres-sures (see Tables I and II).

All pressures returned to initial placebo levels duringthe post-PHN placebo period within 3-6 days after dis-continuing PHN. The patients (F. P. and B. D.) re-ceived a short trial of PHN 1500 mg daily with no fur-

TABLE I IBPs in the Supine Position (mm Hg)

PHN

Pateint Placebo 600 900 Placebo

mg/day

C. H. 142/89 123*/76* 130/77M. L. 160/108 148*/103* 156/103R. XW. 138/94 133/87* 135/87D. K. 164/93 162/98 167/96 181/970. D. 157/101 146*/107 144/99 148/105F. P. 196/138 181*/127* 177*/115* 189/131B. D. 166/105 164/103 168/110 179/109J. S. 177/119 170/112 159*/102* 182/118Mean decrease

from placebo 8.5/5.6 9.0/6.4

All values are means of four measurement's daily for at least4 consecutive days.* Differences between placebo and PHNsignificant P < 0.05with majority P << 0.01.

ther decrease in pressure. Symptoms related to posturalhypotension were not observed during treatment withPHN. Systolic and diastolic BP in the supine positiondecreased significantly in four of the eight patients dur-ing PHN 600 mg daily compared with control (TableII). The mean decrease was 8.5/5.6 mmHg comparedwith placebo. Of the patients who progressed to PHN900 mg daily, only two of five showed a further de-crease in supine pressure. Again, all pressures in thesupine position returned to, or near to, control levelswithin 3-6 days after discontinuing PHN. An exampleof the course of one patient (J. S.) during the adminis-tration of PHN is shown in Fig. 2. There was no sig-nificant change in pulse rate during the study.

The effect of PHN on other indices of sympatheticfunction.. Fig. 3 is representative of the effect ofPHN on the Valsalva maneuver of one patient. Dur-ing PHN, the decrease in pulse pressure was essen-tially unchanged, while the sympathetically mediated

EXPIREDPRESSURE

(mmHg)

ARTERIALPRESSURE

(mmHg)

PLACEBO40-o o.-

20- .

200 - h *[ll tstt

100- .. . ................. ......

P H Nrod1..r t

a ....' . .... . _ . Selili-- tiX!lt;-la

FIGURE 3 Valsalva's maneuver during placebo and PHN.A forced expiratory pressure of 40 mmHg was sustainedfor 10 sec. The changes in the arterial pressure and theirsignificance are described in the text.

954 Rangno, Kaufmann, Cavanaugh, Island, Watson, and Oates

diastolic overshoot was virtually abolished. A coni-posite of the data from Valsalva maneuvers in all pa-tients is shown in Fig. 4. Plots of the pre- and post-PHN data show a normal decrease in pulse pressureand reflex increase in diastolic pressure. During PHN600 mg daily the decrease in pulse pressure was un-changed but reflex rise in diastolic pressure during theovershoot was substantially decreased in all patientsto levels that fall below an established norm (7). Thesechanges are similar to those seen after sympathectomy.

The pressor response to tyramine, an indirectly act-ing amine, and the pressor response to directly actingexogenous NE during placebo and PHN 600 mg dailyare shown in Fig. 5. The pressor sensitivity to tyra-mine decreased markedly in all patients during PHN:that is, a larger dose of tyramine was required to raisethe systolic pressure by 25 mmHg. Tyramine testing wasnot carried out in all subjects because it appeared thattyramine transiently reversed the antihypertensive ef-fects of PHN. The sensitivity to NE increased in fiveout of six patients during PHN; that is, a smaller doseof NE was required to raise the systolic pressure by 25mmHg.

The effect of PHNon the excretion of catecholaininesand catecholamine metabolites. The effect of PHN onthe 24 h urinary VMAis shown in Fig. 6. The progres-sive decrease in urinary excretion of VMAwith increas-ing doses of PHNwas highly significant; the maximumreduction was 57% at PHN 1,500 mg daily.

The effect of PHNon the 24 h urinary NMI and M isshown in Fig. 7. NMI increased about 400% aboveplacebo levels during the administration of PHN900 mgdaily, and returned to normal in post-PHN placebo pe-

* - PRE-PH N0- PHNA - POST- PH N

* ~~~~~-40

\ A

0

A

LEA Aa

U

a

0< --

PERCENTINCREASE

OF

-20 DIASTOLICPRESSURE

80 60 40 20 0

PERCENT DECREASEOF PULSE PRESSURE

FIGURE 4 Results of Valsalva's maneuver in all patientsstudied. Horizontal and vertical axes are calculated accord-ing to Sharpey-Shafer (7). The dotted line represents thelower limit of normal according to Sharpey-Shafer. PHNdose: 600 mg daily during all data shown.

5.74.

3-

2 -

SENSITIVITYRATIO -

0.8

0.6-

0.4

0.2-

TYRAMINE NOREPINEPHRINE

, O.K.

I J.S.l C.H.

, RW

, C.L.

PLACEBO

R.W.OD.K.

IJ.S.

- S.aD.

PHN

I C.L.

C.H.' S.D.

0.1- #PLACEBO PHN

FIGURE 5 Pressor sensitivity ratios of tyramine and nor-epinephrine during placebo and PHN 600 mg daily. Sensi-tivity is expressed as the ratio of the dose of pressoramine required to raise the systolic pressure by 25 mmHgduring placebo compared with the dose required duringadministration of PHN.

riod. MI, in contrast, remained unchanged in all pa-tients studied.

The urinarv excretion of NE is shown in Table IIIin conjunction with that of VMAand NMfor these pa-tients. The decrease in NE excretion was about the samepercentage as the decrease in VMA. In order to evaluatethe effect of PHN on the synthesis of NE, the sum ofVNMA, NM, and NE during PHN 600 mg was com-l)ared with these sumis during the pre- and postplaceboperiods. Total 24 h urinary excretion of the sum of VMA,NM, and NE decreased 23% during PHN administra-tion.

In the four patients studied, urinary DA decreasedsignificantly at all dose levels of PHNwith a maximumdecrease of 50% at PHN 1,500 mg daily, compared withp)re-PHN placebo (Fig. 8). Again, values returned tonear normal levels in the post-PHN placebo period. Theurinary excretion of HVA, the metabolite of DA, wasmeasured in the same four patients. The mean 24 hurine value during pre-PHN placebo was 2.77±0.26mg/g creatinine and did not change significantly duringPHN600 or 900 mg daily or during post-PHN placebo.

The fate of PHN. Substantial oral absorption ofPHNwas verified by the urinary excretion of approxi-mately 60% of the total administered dose as unchangedplus conjugated PHN. This percentage was similar atall dose levels. About 50% of the urinary PHNwas un-

Effects of p-Hydroxynorephedrine in Hypertensive Patients 955

URINARY 3.0 TVMAm

(mg/24 h±SE) 2.0

1 g P001 g~P<0001110- P<0OOOI P<O0I00

0TREATMENT-+PLACEBO PHN PHN PHN PLACEBODAILY DOSE PHN (mg) -+ 600 900 1500

FIGURE 6 Effect of PHN on the 24 h urinary excretionof VMA. Values are the means +SE of consecutive dailycollections during the last 3 days of each period. Of theeight patients, the following numbers completed each treat-ment period: pre- and post-PHN placebo and PHN 600mg: 8/8; PHN900 mg: 5/8; PHN 1500 mg: 2/8. Statis-tical comparison of means is between pre-PHN placebo andeach PHNdose period.

changed and 50% was conjugated. The percentage ofconjugated PHNwas constant at varying doses of PHN.The time-course of urinary excretion of free plus con-jugated PHN was studied in three patients during thepost-PHN placebo period. The first phase of the PHNexcretion had a mean t i of 13 h while the second phaset i had a mean of 55 h (Fig. 9).

The effect of PHNon sodium balance. A negative so-dium balance was observed during PHNin each of fourpatients on a 100 mEqsodium diet, and the single patienton 10 meq sodium diet. Mean values for urinary sodiumexcretion during placebo, PHN600 mg daily, and post-PHN placebo were 89.0, 114.4, and 82.6 meq/24 h, re-

spectively. There was a 28% reduction in sodium excre-

tion from the period of negative sodium balance duringPHN to that of positive balance during post-PHNplacebo (P < 0.02). The same mean values for urinarysodium for the single patient (F. P.) on a 10 meq so-

dium diet were 17.8, 19.1, and 4.6 meq/24 h. The positivebalance from PHN to post-PHN placebo was again sig-nificant (P < 0.01).

Other evaluations. There were no significant changesin the following routine tests throughout this study:hemoglobin, hematocrit, white blood cell count, differ-ential, platelets, fasting blood sugar, blood urea nitrogen,total bilirubin, serum glutamic oxaloacetic transaminase,uric acid, total proteins, alkaline phosphatase, serum so-

dium, potassium, chloride and bicarbonate, and uri-nalysis.

There was no evidence of psychic depression or stimu-lation at any level of PHN dosage. These observationswere verified by a blinded psychiatric evaluation. Acontinuing review of systems failed to elicit any otheruntoward effects, except for the possibility of mild con-stipation in one female. Sexual function was difficultto assess in these hospitalized patients but direct ques-tioning suggested no change in libido and two of themale patients on short leave during PHNadministrationdescribed normal potency.

With the exception of one patient, there was no sig-nificant pressor effect during initial dosing with PHN100 mg. The change in mean systolic pressure from timezero to i, 1, and 11 h post-PHN (or placebo) was + 6,+ 7, and + 5 mmHg during placebo, and - 2.4, + 6, and+ 5 mmHg during PHN. Single large doses of PHNin excess of 100 mg produced a mild transient increasein BP of 30/5 mmHg in one patient when PHNwas ini-tiated at this dose level. After 3 days, doses in excess of100 mg did not cause a pressor change in this patient.Single doses as large as 300 mg were tolerated by otherpatients without pressor effect; however, initial doses inthese patients were never in excess of 100 mg.

An abnormality of ECG occurred in one patient(F. P.) during PHN. This patient had the most severe

TREATMENT PLACEBO PHN PHN PHNDAILY DOSE(mg) -- 600 900 1500

FIGURE 7 Effect of PHNon the 24 h urinary excretion ofNMand M. Values are the means +SE of at least threecollections during the end of each treatment period. Of thethree patients studied, the following numbers completedeach treatment period: pre- and post-PHN placebo andPHN 600 mg: 3/3; PHN 900 mg; 2/3; PHN 1500 mg:1/3.

956 Rangno, Kaufmann, Cavanaugh, Island, Watson, and Oates

hypertension, and required a 10 meq sodium diet throughthe placebo and PHN periods. She described chronicpalpitations before and after the study. Before the studyand during pre-PHN placebo, ECGs revealed left ven-tricular hypertrophy and first degree heart block as wellas an intraventricular conduction defect. During PHN600 and 900 mg daily, there was evolution of the coni-duction defect to right bundle branch block (RBBB).The patient noted palpitations but all vital signs werestable. Serial lactic dehydrogenase and serum glutamicoxaloacetic transaminase were unchanged. When PHNwas discontinued, the RBBBreverted to an intraventricu-lar conduction defect. At the time of her discharge, herhypertension was treated with a combination of hydro-chlorothiazide, guanethidine, and methyldopa. RepeatedECGs during the post discharge period have shown pe-riodic RBBB.

DISCUSSIONAlthough PHN is a minor metabolite of amphetamine,its formation after amphetamine administration occursalmost exclusively within adrenergic neurons; the pri-mary metabolite of amphetamine, PHA, is actively trans-ported into the neuron by the NE pump (14) and is9-hydroxylated intraneuronally (15). The administra-tion of PHN to patients produces inhibition of sympa-thetic reflexes, as indicated by almost complete abolitionof the pressor overshoot of the Valsalva reflex. Thesefindings support earlier suggestions that alterations insympathetic function produced by d-amphetamine (1)and p-hydroxyamphetamine (4) could be mediated bytheir active metabolite, PHN.

It is most likely that sympathetic reflexes are inhibitedprimarily at the level of the peripheral adrenergic neu-

TABLE I IIUrinary Catecholamines during PHNCompared

with Pre- and Post-PHN Placebo

24-h urinarycatecholamines* PHN(jg/g creatinine) Placebo (600 mg/day) Placebo

VMA 3235±259 1932±181T 3123±274NM 277+24 773±81: 267±37NE 61±7 35±6: 48±3

Total 3573 2740§ 3438

VMA, vanillylmandelic acid; NM, normetanephrine; NE,norepinephrine.* Values are the means 45E of three separate 24-h urines ineach period of treatment in the two patients in whomall thesemeasurements were made (i.e., n = 6).4 Differences between placebo and PHN significant(P < 0.01).§ 23% decrease.

300-

URINARYDOPAMINE

CREATI NI NE± SE

100-

P<o-0o11 P< aoo1

0'VTREATIMIENT-4'PLACEBO PHN PHN PHN PLACEBODAILY DOSEPHN (mg) -.600 900 1500

FIGURE 8 Effect of PHN on the 24 h urinary excretionof DA. Values are the means ±SE of three collections dur-ing the each of each treatment period. Of the four patientsstudied, the following numbers completed each treatmentperiod: pre- and post-PHN placebo and PHN 600 mg:4/4; PHN 900 mg: 3/4; PHN 1500 mg: 2/4.

ron. A polar amine such as PHN should not readilycross the blood-brain barrier. Recent studies in ourlaboratory have shown catecholamines were depletedfrom adrenergic neurons in the rat heart by PHN, whilebrain catecholamine content was little altered. This is in

500-

1st PHASE

MEANI 2 = 13 h100 - la

URINARYPHN

mg/24h \

10 0 £

2nd PHASEMEANt 55 h

0 2 4 6 8 10 12TIME AFTER PHN STOPPED(day)

FIGURE 9 Delayed urinary excretion of PHN in threepatients discontinuing PHN.

Effects of p-Hydroxynorephedrine in Hypertensive Patients 957

keeping with our clinical observations showing littlediscernible effect of PHN on the psychic function ofthe patients in this study. This is in marked contrastto the effects seen when false transmitters can be de-livered into the brain in the form of an amino acid pre-cursor such as methyldopa. Whereas a substitutedphenylethylamine such as PHNmay not easily enter thebrain, it can be readily transported into the peripheraladrenergic neuron by the NE pump (14). Within theneuron, PHNcan interfere with adrenergic function inat least two ways.

There is considerable evidence that it can replace NEwithin the neurosecetory vesicle from which it, in turn,can be released as a less potent "false neurotransmitter"(2, 3, 16). Another false neurotransmitter, metaraminol(m-hydroxynorephedrine) has also shown to have anti-hypertensive properties in man (17).

In our patients, NEdepletion by PHNfrom the periph-eral adrenergic neurons is suggested by the marked de-crease in the pressor sensitivity to tyramine. The inhibi-tion of tyramine's effect must be at the neuronal levelbecause the slightly increased pressor response to ex-ogenous NE indicates that the response of the receptoris not decreased. Part of the diminution in tyramine ef-fect and the slight enhancement of NE effect could bedue to competitive inhibition of tyramine uptake intothe neuron by PHN. It is of interest that the diminishedresponse to tyramine seen with PHN is in contrast tothe enhanced response seen when the more potent alphaadrenergic agonist, alpha methylnorepinephrine, is thefalse transmitter (18).

In addition to replacement of NE within the secretorystores, the excretion of a diminished amount of catechola-mines and total catecholamine metabolites suggests thatthe false neurotransmitter, PHN, also inhibited NE syn-thesis in these patients. The 23% decrease in the totalmeasured metabolites of catecholamines is probably anunderestimation of the decrease in NE synthesis becausea fall in excretion of VMAin the presence of increasedexcretion of NM implies that the formation of theoxidative metabolite of the catecholamines, dihydroxy-mandellic acid, was also decreased. The decrease in ex-cretion of both catecholamines and their metabolites isall the more remarkable in light of the concomitant fallin BP which alone is a stimulus for increased sympa-thetic activity and increased NE synthesis. These findingswith PHNin man are in line with the observation thatthe related false transmitters, alpha-methylnorepinephrineand m-hydroxynorephedrine, cause a decrease in NE syn-thesis in the rat, as reported by Kopin, Weise, and Sed-vall during the course of these studies (19). Because DAexcretion was decreased and HVAexcretion unchanged,it is most likely that the decrease in NE synthesis occursbefore the beta hydroxylation step. A likely mechanism

would be release of NE from neurosecretory vesicles intothe cytoplasm of the neuron where an increase in NEconcentration would inhibit tyrosine hydroxylase (20,21).

Because normetanephrine constitutes only a small frac-tion of total NE metabolites, total measured metabolitesfell even though NMs excretion rose. This observed in-crease in catecholamine o-methylation would occur if NEuptake into the neuron (and exposure to intraneuronaloxidation) were blocked. Some inhibition of NE reup-take due to competitive inhibition of the NE pump byPHNis likely. There are similar changes in NMexcre-tion when NE reuptake is blocked by imipramine (22).In light of the decrease in the pressor response to tyra-mine, partial inhibition of the NE pump is a more likelyexplanation of the increased normetanephrine excretionthan is monoamine oxidase inhibition.

The observed interference with sympathetic controlof arteriolar tone by PHNmust be a major factor in thereduction of BP seen in these hypertensive patients inwhom the fall in BP was greater in the upright than inthe supine position. The increased excretion of sodiummay also contribute to the antihypertensive effect, andis in contrast to the effect of guanethidine which causessodium retention (23).

In these patients, the hypotensive dose range was300-900 mg daily. The mild transient pressor effectnoted in one patient when the dose was rapidly pro-gressed to single doses of 100 mg or greater probablyrepresents an indirect pressor effect; it did not occur ona second trial of the drug during which doses of PHNlarger than 100 mg had no pressor effect after initialexposure to lower doses.

The urinary excretion of PHNis a biexponential func-tion. The first phase (t A 13 h) represents excretion froma pool of relatively large mass, while the second phase(t i 55 h) represents excretion from a smaller poolwhich is more avidly retained at its site of extravascu-lar storage. The time-course of disposition of PHNfromthis second pool closely parallels the disappearance ofthe hypotensive effect and the reversal of the changes incatecholamine synthesis and metabolism. This correla-tion with the effect on adrenergic neurons, the analogywith the neuronal storage of guanethidine (24), and theevidence for a neuronal pool of m-hydroxynorephedrine(25) all suggest that the slowly released pool repre-sents that PHN stored in adrenergic neurons. The pro-longed retention of this amine intraneuronally is fa-cilitated by the alpha methyl group which confers rela-tive protection from degradation by monoamine oxidase.The long half-life of the slowly released pool indicatesthat PHNwill accumulate in it during continue adminis-tration of either amphetamine or PHN itself. Assumingthat near maximal accumulation occurs after administra-

958 Rangno, Kaufmann, Cavanaugh, Island, Watson, and Qates

tion of PHN for about five half-lives, then its accumu-Iation in this slowly released pool would continue for11.5 days, reaching 87.5% of maximum levels after 7days. When its precursor, amphetamine, is ingested on acontinuing basis, maximal accumulation of PHN would-not occur until 11.5 days after the several days requiredfor maximal accumulation of amphetamine.

During amphetamine ingestion, the selective synthe-sis and storage of PHN in the adrenergic neuron and itsaccumulation to a greater extent than the parent drugsuggests that a pharmacologic effect of this minor activemetabolite would become maximal when amphetamineis ingested over many days as occurs with the chronicabuse or prolonged prescribed use of amphetamine. In-deed, in our earlier studies on short term ingestion ofamphetamine, evidence for the effect of a false transmitter(e.g., decreased pressor response to tyramine) was notapparent until amphetamine had been taken for more thana day and even at 5 days there did not appear to bemaximal inhibition of the pressor response to tyramine(1). These finds and the data on the half-life of PHN-predict that the more extended ingestion of amphetamineshould produce evidence of the effect of a false trans-mitter that is even more profound than observed in ourearlier study on the short term ingestion of amphetamine.An evaluation of adrenergic function by measures suchthe pressor response to tyramine and the Valsalva ma-neuver in chronic amphetamine abusers would be of in-terest and could lead to a rapid diagnostic test for thiscondition.

When amphetamine is administered repeatedly, thereis tachyphylaxis to its pressor effect (26). This has beenconfirmed in a recent study in animals which concludedthat PHN might be involved in the tolerance to theperipheral but probably not to the central effects of am-phetamine (27). This accounts for the fact that abusersof the drug can escalate the dose of amphetamine toamounts that would produce disastrous cardiovasculareffects if taken as an initial dose. The diminished pres-sor response to another indirectly acting amine, tyra-mine, during continuing administration of either PHNor amphetamine suggests that depletion of NE by PHNis responsible for the tachyphylaxis of cardiovasculareffects to amphetamine.

In summary, this investigation establishes that aperipherally acting false neurotransmitter such as PHNcan interfere with adrenergic neuron function, an effectthat is associated in hypertensive patients with a re-duction in BP. The lack of central side effects is notable,and suggests that a compound capable of acting as aperipheral false transmitter might have therapeutic po-tential if the ratio of its pressor to catecholamine de-pleting effect were appropriately low.

ACKNOWLEDGMENTSThe authors wish to acknowledge the capable technical as-sistance of Mr. Keith Carr, Mrs. Elizabeth Winkert, Mrs.Becky Bilbrey, and Mrs. Ellen Nuckolls.

This work was supported by grants HE 10842, HE 05545,and research fellowship award GM23168 to Dr. J. Kauf-mann.

REFERENCES1. Cavanaugh, J. H., J. D Griffith, and J. A. Oates. 1970.

Effect of amphetamine on the pressor response to tyra-mine: formation of p-hydroxynorephedrine from am-phetamine in man. Clin. Pharmacol. Ther. 11: 656.

2. Daly, J. W., C. R. Creveling, and B. Witkop. 1966. Thechemorelease of norepinephrine from mouse hearts.Structure-activity relationships. I. Sympathomimetic andrelated amines. J. Med. Chem. 9: 273.

3. Thoenen, H., A. Hfirlimann, K. F. Gey, and WV. Haefely.1966. Liberation of p-hydroxynorephedrine from catspleen by sympathetic nerve stimulation after pretreat-ment with amphetamine. Life Sci. 5: 1715.

4. Gill, J. R., Jr., D. T. Mason, and F. C. Bartter. 1966.Effects of hydroxy-amphetamine (Paredrine) on thefunction of the sympathetic nervous system in normo-tensive subjects. J. Pharinacol. Exp. Ther. 155: 288.

5. Ohler, E. A., and G. E. Wakerlin. 1953. Amines inexperimental hypertensions. Circ. Res. 1: 122.

6. Vogin, E. E. 1968 Report: acute and subacute safetyevaluations of para-hydroxynorephedrine in rats anddogs. Food and Drug Research Laboratories, Maspeth,N. Y. Lab. no. 88893.

7. Sharpey-Schafer, E. P. 1955. Effects of Valsalva'smaneuver on the normal and failing circulation. Br.Med. J. 1: 693.

8. Pisano, J. J., J. R. Crout, and D. Abraham. 1962. De-termination of 3-methoxy-4-hydroxymandelic acid inurine. Clint. Chini. Acta. 7: 285.

9. Crout, J. R. 1961. Catechol amines in urine. Stanid.Methods Cliii. Chew. 3: 62.

10. Anton, A. H., and D. F. Sayre. 1966. Distribution ofmetanephrine and normetanephrine in various animalsand their analysis in diverse biologic material. J. Phar-macol. Exp. Ther. 153: 15.

11. Anton, A. H., and D. F. Sayre. 1964. The distributionof dopamine and dopa in various animals and a methodfor their determination in diverse biological material.J. Pharmacol. Exp. Ther. 145: 326.

12. Sato, T. L. 1965. The quantitative determination of 3-methoxy-4-hydroxyphenylacetic acid (homovanillic acid)in urine. J. Lab. Clin. Mcd. 66: 517.

13. Pisano, J. J., J. A. Oates, Jr., A. Karmen, A. Sjoerdsma,and S. Udenfriend. 1961. Identification of p-hydroxy-a-(methylaminomethyl)benzyl alcohol (Synephrine) inhuman urine. J. Biol. Chew. 236: 898.

14. Burgen, A. S. V., and L. L. Iversen. 1965. The inhibi-tion of noradrenaline uptake by sympathomimetic aminesin the rat isolated heart. Br. J. Pharmctacol. 25: 34.

15. Sjoerdsma, A., and W. von Studnitz. 1963. Dopamine-#-oxidase activity in man, using hydroxyamphetamineas substrate. Br. J. Pharnmacol. 20: 278.

16. Chen, K. K., C. Wu, and E. Henriksen. 1929. Relation-ship between the pharmacological action and the chemi-cal constitution and configuration of the optical isomersof ephedrine and related compounds. J. Pharmacol. Exp.Ther. 36: 363, 394. (Addendum)

Effects of p-Hydroxynorephedrine in Hypertensive Patients 959

17. Cohen, R. A., I. J. Kopin, C. R. Creveling, J. M. Musac-chio, J. E. Fisher, J. R. Crout, and J. R. Gill, Jr. 1966.False neurochemical transmitters. Ann. Intern. Med.65: 347.

18. Pettinger, W. A., D. Horwitz, S. Spector, and A.Sjoerdsma. 1963. Enhancement by methyldopa of tyra-mine sensitivity in man. Nature (Lond.). 200: 1107.

19. Kopin, I. J., V. K. Weise, and G. C. Sedvall. 1969. Effectof false neurotransmitters on norepinephrine synthesis.J. Pharmacol. Exp. Ther. 170: 246.

20. Ikeda, M., L. A. Fahien, and S. Udenfriend. 1966. Akinetic study of bovine adrenal tyrosine hydroxylase.J. Biol. Chem. 241: 4452.

21. Alousi, A., and N. Weiner. 1966. The regulation ofnorepinephrine synthesis in sympathetic nerves: effect ofnerve stimulation, cocaine, and catecholamine-releasingagents. Proc. Natl. Acad. Sci. U. S. A. 56: 1491.

22. Schildkraut, J. J., E. K. Gordon, and J. Durell. 1965.Catecholamine metabolism in affective disorders. I. Nor-

metanephrine and VMAexcretion -in depressed patientstreated with imipramine. J. Psychiatr. Rev. 3: 213.

23. Jose, A., J. R. Crout, and N. M. Kaplan. 1970. Sup-pressed plasma renin activity in essential hypertension.Roles of plasma volume, blood pressure, and sympa-thetic nervous system. Ann. Intern. Med. 72: 9.

24. Oates, J. A., J. R. Mitchell, 0. T. Feagin, J. S. Kauf-mann, and D. G. Shand. 1971. Distribution of guanidin-ium antihypertensives: mechanism of their selective ac-tion. Ann. N. Y. Acad. Sci. 179: 302.

25. Shore, P. A., D. Busfield, and H. S. Alpers. 1964. Bind-ing and release of metaraminol: mechanism of norepi-nephrine depletion by a-methyl-m-tyrosine and relatedagents. J. Pharmacol. Exp. Ther. 146: 194.

26. Goodman, L. S., and A. Gilman. 1967. The Pharma-cologic Basis of Therapeutics. The MacMillan Com-pany, New York. 3rd edition. 298.

27. Lewander, T. 1971. A mechanism for the developmentof tolerance to amphetamine in rats. Psychopharmacolo-gia. 21: 17.

960 Rangno, Kaufmann, Cavanaugh, Island, Watson, and Oates